Prevalence of Substance
VOL. 16, NO. 1, 1990
by Kim T. Mueser, Paul R.
Yarnold, Douglas F. Levinson,
Hardeep Singh, Alan S. Bellack,
Kimmy Kee, Randall L. Morrison,
and Kashinath G. Yadalam
Abuse in Schizophrenia:
Demographic and
Clinical Correlates
Abstract
Methodological issues involved in
assessing the prevalence of sub-
stance abuse in schizophrenia are
discussed, and previous research in
this area is comprehensively re-
viewed. Many studies suffer from
methodological shortcomings, in-
cluding the lack of diagnostic rigor,
adequate sample sizes, and simul-
taneous assessment of different
types of substance abuse (e.g.,
stimulants, sedatives). In general,
the evidence suggests that the
prevalence of substance abuse in
schizophrenia is comparable to that
in the general population, with the
possible exceptions of stimulant
and hallucinogen abuse, which may
be greater in patients with schizo-
phrenia. Data are presented on the
association of substance abuse with
demographics, diagnosis, history of
illness, and symptoms in 149 recent-
ly hospitalized DSM-III-R schizo-
phrenic, schizophreniform, and
schizoaffective disorder patients.
Demographic characteristics were
strong predictors of substance
abuse, with gender, age, race, and
socioeconomic status being most
important. Stimulant abusers tend-
ed to have their first hospitalization
at an earlier age and were more
often diagnosed as having schizo-
phrenia, but did not differ in their
symptoms from nonabusers. A
history of cannabis abuse was
related to fewer symptoms and
previous hospitalizations, sug-
gesting that more socially compe-
tent patients were prone to cannabis
use. The findings show that en-
vironmental factors may be impor-
tant determinants of substance
abuse among schizophrenic-
spectrum patients and that clinical
differences related to abuse vary
with different types of drugs.
31
The prevalence of substance abuse
in schizophrenia, as well as its in-
fluence on the etiology and course
of the disorder, is an important but
unstudied problem in clinical
psychiatry. A wide range of different
substances produce symptoms that
mimic schizophrenia, but the in-
fluence of substance abuse on
schizophrenia is controversial.
Chronic alcohol abuse and with-
drawal from alcohol can produce
psychotic symptoms, including delu-
sions and hallucinations (Victor and
Hope 1958; Mott et al. 1965; Schuckit
1989), and have been hypothesized
both to hasten the onset of schizo-
phrenia (Freed 1975) and to mask its
presence (Diethelm 1957; Parker et
al. 1960; DeVito et al. 1970). Some
schizophrenic patients report that
drinking reduces their symptoms
(Alpert and Silvers 1970; Hansell
and Willis 1977), while others report
the opposite (Kesselman et al. 1982;
Schuckit 1983). Cannabis abuse can
induce brief paranoid reactions,
panic attacks, and mental confusion
in persons with no prior psychiatric
illness (Clark et al. 1970; Chopra and
Smith 1974; Hollister 1986), and it
has been reported to cause symptom
exacerbations in schizophrenic sub-
jects (Bernhardson and Gunne 1972;
Treffert 1978; Knudsen and Vilmar
1984). According to medical lore,
chronic cannabis use can lead to a
psychotic state resembling schizo-
phrenia (Chopra 1971; Thacore
1973), although the validity of this
phenomenon has been questioned
(Thacore and Shukla 1976; Ghodse
1986; Hollister 1986). Acute and
chronic amphetamine challenge can
Reprint requests should be sent to Dr.
K.T. Mueser, The Medical College of
Pennsylvania at EPPI, 3200 Henry Ave.,
Philadelphia, PA 19129.
32
cause paranoid ideation and halluci-
nations that can be indistinguishable
from schizophrenia in previously
unimpaired persons (Bell 1965; Hall
et al. 1988). Amphetamine abuse
has repeatedly been shown to
worsen symptoms in some (Jan-
owsky et al. 1973; West 1974; Jan-
owsky and Davis 1976) but not all
schizophrenic patients (van Kam-
men et al. 1982, 1985; Angrist et al.
1985; Lieberman et al. 1987). In
addition, clinicians have reported
successfully treating schizophrenic
patients solely or adjunctively with
amphetamines, especially patients
with prominent negative symptoms
(Wooley 1938; Davidoff and Reifen-
stein 1939; Angrist et al. 1980, 1982;
Cesarec and Nyman 1985).
In the 1960's and 1970's, the youth
counterculture revived the popular-
ity of several psychoactive sub-
stances that had enjoyed wider use
earlier in the century, including can-
nabis, amphetamines, cocaine, and
narcotics. Hallucinogens such as
d-lysergic acid diethylamide (LSD),
methylene dioxymethamphetamine
(MDA), mescaline, and psilocybin
were also introduced into the recrea-
tional drug marketplace. The use of
these substances was associated
with psychopathological reactions
including brief psychotic episodes
(Hensala et al. 1967; Freedman
1968), schizophreniform psychoses
(Bowers 1972, 1977; Bowers and
Swigar 1983; Vardy and Kay 1983),
and suicides (Cohen 1964, 1966).
The apparent increase in sub-
stance abuse by psychiatric patients
has been particularly prominent
among young patients with chronic
mental illness, the majority of whom
have schizophrenia (Pepper et al.
1981; Safer 1987). This is of special
concern since substance abuse
among young patients has been
found to worsen the course of ill-
ness, increasing their risk of symp-
tom exacerbations and rehospitaliza-
tions (Crowley et al. 1974; Carpenter
et al. 1985; Safer 1987; Drake et al.
1989). To complicate matters further,
"dual-diagnosis" patients tend to
receive less outpatient treatment and
to be more difficult to treat in the
community, resulting in a poor
prognosis (Solomon 1986; Solomon
and Davis 1986).
Prevalence of Substance
Abuse
To examine research on the preva-
lence of substance abuse in
schizophrenia, a comprehensive
methodological review of published
studies in this area was conducted.
The scope of the review was limited
to studies examining either alcohol
or illicit drug abuse in schizophre-
nia, excluding studies of caffeine or
tobacco use (for a review, see
Schneier and Siris 1987). Studies
were included if they had at least 15
patients and if the specific number
of patients in each diagnostic group
who were and were not substance
abusers was reported (or could be
calculated). Finally, only those
studies were included which
examined abuse in a sample of pa-
tients who were not selected on the
basis of a history of abuse (or lack
thereof). Thus, studies comparing a
fixed number of substance-abusing
patients with a group of nonabusing
patients drawn from a different sam-
ple were excluded, as were studies
that failed to report the total number
of patients assessed or those that
focused solely on characteristics of
substance-abusing psychiatric pa-
tients (e.g., Roy 1981; Vardy and Kay
1983; Hays and Aidroos 1986; Ross
et al. 1988; Pulver et al. 1989).
Table 1 summarizes the research
SCHIZOPHRENIA BULLETIN
conducted on substance abuse in
schizophrenia, including the
methodological characteristics of
each study. Studies on only alcohol
are presented first, followed by
studies examining specific classes of
drug abuse (e.g., stimulants, can-
nabis), and finally by studies on
unspecified classes of substance
abuse. The relevance of each of the
listed methodological issues to
estimations of the prevalence of
substance abuse in schizophrenia is
discussed below.
Diagnosis of Patients. Accurate,
reliable psychiatric diagnoses based
on specific, widely accepted criteria
are essential in assessing the preva-
lence of substance abuse in psychi-
atric populations. Diagnostic
systems developed before the criteria
of Feighner et al. (1972) had poor
reliability (Spitzer and Fleiss 1974).
Even with the use of operationally
defined criteria (e.g., the Research
Diagnostic Criteria [RDC] of Spitzer
et al. [1978]), schizophrenia was less
reliably diagnosed than other major
psychiatric disorders (Helzer et al.
1977). Accurate diagnosis is critical
to evaluating the psychopathological
consequences of commonly abused
substances such as alcohol, stimu-
lants, and hallucinogens. A second
aspect of diagnostic assessment is
whether a structured interview was
used to obtain information on symp-
toms and history of the illness. The
failure to use a structured diagnostic
interview increases information
variance and the risk of misdiag-
nosis significantly (Alterman et al.
1984).
More than one-third (8 out of 22)
of the studies reviewed did not
specify which diagnostic criteria
were used (table 1). Only three
studies reported using standardized
structured interviews in diagnosing
VOL. 16, NO. 1, 1990
patients (Siris et al. 1988; Ananth et
al. 1989; Barbee et al. 1989).
Subject Characteristics. To evaluate
whether the prevalence of substance
abuse in schizophrenia differs from
that in other psychiatric disorders or
the general population, it is desir-
able to have a comparison group of
subjects that is matched to the
schizophrenic group on demo-
graphic characteristics. Gender, age,
socioeconomic status, and race have
all been found to be important fac-
tors of substance abuse in the
general population (e.g., Smith
1989). If these variables are not con-
trolled for when comparing diag-
nostic groups, differences in
prevalence rates could be errone-
ously attributed to diagnostic rather
than demographic factors.
Gender is a particularly important
factor for both substance abuse and
schizophrenia. There is overwhelm-
ing evidence that males are more
prone to substance-abuse disorders
than are females in the general
population (Myers et al. 1984; Smith
1989). It has long been observed that
females with schizophrenia tend to
have a more benign course of the ill-
ness (Angermeyer and Kuhn 1988)
and spend less time in psychiatric
hospitals than males (Goldstein
1988). A result of these gender dif-
ferences is that relatively more male
than female schizophrenic subjects
are treated in inpatient settings,
where most studies of substance-
abuse prevalence are conducted. A
high male-to-female ratio in a sam-
ple of schizophrenic patients could
result in higher estimates of
substance abuse for this disorder.
Seven of the studies reviewed in-
cluded only males or predominantly
males, seven studies specified the
number of male and female schizo-
phrenic patients in the sample, and
eight studies provided no informa-
tion on the gender distribution of
the schizophrenic sample. However,
only three of the seven studies
reporting the number of male and
female schizophrenic patients pre-
sented data on substance abuse ac-
cording to gender (Whitlock and
Lowrey 1967; Negrete et al. 1986;
Barbee et al. 1989). Few studies have
examined the relationships between
substance abuse and other demo-
graphic characteristics in schizo-
phrenic patients (e.g., age, race, and
socioeconomic status). These need
to be explored to determine whether
the same demographic factors that
predispose nonpsychiatric patients
to substance abuse are also associ-
ated with greater abuse among
schizophrenic patients.
Hospital Setting. The setting where
the study sample is obtained may
have an important bearing on
substance abuse prevalence esti-
mates (Galanter et al. 1988).
Estimates of the prevalence of abuse
in schizophrenic patients requiring
emergency treatment may be higher
than estimates obtained from other
settings (Atkinson 1973; Barbee et al.
1989) as has been found in nonpsy-
chiatric patients requiring emer-
gency treatment (e.g., Trier and
Levy 1969; Atkinson 1973). The
prevalence of substance abuse may
also vary depending on whether the
patients are inpatients or outpatients
and on the chronicity of illness.
Some research has suggested that
more severely ill psychiatric patients
are less likely to be substance
abusers (or "heavy" abusers) than
less ill patients (Cohen and Klein
1970; Ritzier et al. 1977; OTarrell et
al. 1983). However, this suggestion is
not supported by the differences in
abuse between acute and chronic
inpatients found in the studies
33
reviewed (table 1). Similarly, while
schizophrenic patients in the
Veterans Administration (VA) system
tend to have better premorbid func-
tioning than State hospital patients
(Zigler and Levine 1973), there are
no clear differences in abuse patterns
between VA patients (Pokorny 1965;
McLellan and Druley 1977; Alterman
et al. 1981; OTarrell et al. 1983;
Magliozzi et al. 1983) and others.
Definition of Substance Abuse. Ac-
curate measurement of substance
abuse is a difficult problem for all
populations (Nirenberg and Maisto
1987; Donovon and Marlatt 1988).
Definitions of substance abuse vary
from simple "use" of a psychoactive
substance to "abuse" or "depen-
dence." Abuse generally refers to the
repeated use of a substance to the
extent that it interferes with ade-
quate social, vocational, or self-care
functioning. Dependence refers to
the development of tolerance to a
substance such that the person re-
quires larger dosages to achieve the
same psychoactive effect, and the
experience of withdrawal symptoms
and craving after a period of
abstinence from the substance
(DSM-III-R; American Psychiatric
Association 1987). In practice,
substance abuse and dependence
are often difficult to distinguish
from each other, particularly for can-
nabis and hallucinogen use, for
which tolerance and a clear pattern
of withdrawal symptoms have not
been established. With psychiatric
patients, whose everyday function-
ing is impaired by their illness, the
distinction between substance use
and abuse is also difficult to make,
since the relative influence of
psychiatric illness and substance
abuse on current functioning is
cloudy.
34 SCHIZOPHRENIA BULLETIN

Table 1. Epidemiological studies of substance abuse in schizophrenia

Diagnostic i Hospital3 Definition4 Duration5

2
Study criteria Subjects setting of abuse of abuse Alcohol

Alcohol only studies

Parker et al. Lewis & S:150(M) Al A Life S:22.0
(1960) Piotrowski MD:70(M) MD:32.8
(1954) D:100(M) D:20.0

Pokorny ? S:89(M)7 Al/ A Life S:15.7

(1965) MD:20(M) VA MD:15.0
P:44(M) P:19.0
N:84(M) N:25.0
OBS:18(M) OBS:0

Alterrnan et al. DSM-II S:578(M) Cl/ A Life S.12.3

(1981) OBS:169(M) VA OBS:21.9
O:120(M) 0:19.2

O'Farrell et al. ? S:2078 Cl/ A Life S:23.0

(1983) A:37 VA A:63.0
OBS:40 OBS:35.0

Bernadt & RDC S:57 Al A/ Past S:8.8

Murray SA:16 MHDS year SA:6.3
(1986) D:34 D:20.6
M:30 M:20.0
MJD:69 MJD:11.6
0:86 0:7.0
OC:16 OC:6.3

Alcohol/drug abuse studies

Rockwell & ? S:86(M)9 Cl A10 Current
Ostwald
(1968)

Cohen & Klein S:24 Al A,D Life

Klein (1967) P:67
(1970)

See footnotes at end of table.

VOL. 16, NO. 1, 1990 35

Percentage of subjects with substance abuse6

Stimulant Sedative Cannabis Hallucinogen Narcotics Unspecified

S:0[A]

S:12.5 S:33.3
(Abuse) (Abuse)
P:10.4 P:26.9
(Abuse) (Abuse)
S:8.3
(Dependence)
P:49.2
(Dependence)
 36 SCHIZOPHRENIA BULLETIN

Table 1. Epidemiological studies of substance abuse in schizophreniaContinued

Diagnostic1 Hospital3 Definition4 Duration5

2
Study criteria Subjects setting of abuse of abuse Alcohol
Breakey et al. Breakey & S:28(M)" Al A Life
(1974) Goodell S:18(F)
(1972) C:28(M)
C:18(F)

Hansell & Taylor & S:276(M) 0 A12

Willis Abrams S:299(F)
(1977) (1975)

McLellan & ? S:141(M) \l A Life S:14.2

Druley D:87(M) VA D:17.2
(1977) O:60(M) 0:15.0

Magliozzi DSM-III S:57(M) 1/ (J13 Current

et al. VA
(1983)

Richard S:141 Al u Past 6

et al. A:55 months
M9851 P:19

Negrete ICD-9 S:82(M) O U14 Life,

et al. S:55(F) past 6
(1986) months

Siris et al. RDC15/ S/SA:24(M) O A Life

(1988) SADS S/SA:22(F)

Barbee et al. DSM-II/ S:35(M) ER A,D/ Life S:31.4(M)16

(1989) DIS S:18(F) DIS (Abuse)
S:27.8(F)
(Abuse)
S:42.8(M)
(Dependence)
S:27.8(F)
(Dependence)
See footnotes at end of table.
VOL. 16, NO. 1, 1990 37

6
Percentage of subjects with substance abuse
Stimulant Sedative Cannabis Hallucinogen Narcotics Unspecified
S:15.0[A] S:45.0[M] S:20.0[L]
C:6.5[A] C:15.2[M] C:8.7[L]
S:17.5[H] S:2.5[ME]
C:10.9 H] C:10.9[ME]
S:4.5[A]

S:11.3[A] S:3.5[B] S:9.9 S:6.4

D:2.3[A] D:13.8[B] D:1.1 D:9.2
0:11.7[A] O:10.0[B] 0:11.7 0:8.3

S:42.0
A:19.0
PTS:25.0

S:23.1
A:3.6
P:15.8
S:15.4[C]
A:1.8[C]
P:O[C]
S:65.8(M)
(Life)
S:40.0(F)
(Life)
S:24.0(M)
(6 months)
S:9.1(F)
(6 months)

S:SA:13.0[A] S/SA:4.0 S/SA:35.0[M] S/SA:11.0 S/SA:2.0

S:SA:13.0[C]

S:8.6[B] S:35.8[M] S:5.7 S:6.9

S:3.4[C] S:1
38 SCHIZOPHRENIA BULLETIN

Table 1. Epidemiological studies of substance abuse in schizophreniaContinued

Diagnostic1 Hospital3 Definition4 Duration5

2
Study criteria Subjects setting of abuse of abuse Alcohol

Drake et al. DSM-II S:68(M) o U.A.D/ Past 6 S:7.017

(1989) S:47(F) CPS months (Abuse)
S:14.8
(Dependence)
Unspecified substance abuse studies

Whitlock & ? S:39(M)19 Al D Past

Lowrey S:44(F) month
(1967) MD:14(M)
MD:45(F)
D:24(M)
D:57(F)

Atkinson ? S:89 ER A Current

(1973) A:20
P:42
N:52
OBS.23
C:13

Hall et al. RDC, S:52 O A21 Current

(1977) NHSI A:60
P:27
N:43

Bowers & DSM-III S:17 Al U22 Past 3

Swigar SA:20 years
(1983) SP:21
M:24

Safer ? S:35 O A Before 3

(1987} 0:33 months ago
(past)
Past 3 months
(recent)

See footnotes at end of table.

VOL. 16, NO. 1, 1990 39

Percentage of subjects with substance abuse6

Stimulant Sedative Cannabis Hallucinogen Narcotics Unspecified

S:33.918

S:2.6(M)
S:6.8(F)
MD:7.1(M)
MD:17.7(F)
D:4.2(M)
D:26.3(F)

S:29.22o
A:20.0
P:40.5
N:40.4
OBS:17.4
C:30.8

S:15.4
A:16.7
P:11.1
N:20.9

S:64.7
SA:70.0
SP:66.7
M:54.0

S:42.8
(Past)
0:21.2
(Past)
S:40.0
(Recent)
0:12.1
(Recent)
40 SCHIZOPHRENIA BULLETIN

Table 1. Epidemiological studies of substance abuse in schizophreniaContinued

Diagnostic1 Hospital3 Definition4 Duration5

2
Study criteria Subjects setting of abuse of abuse Alcohol

Ananth et al. DSM-III/ S:38 Al A,D/ 23 Life

(1989) DIS A:17 DIS
O:20

diagnostic criteria used/structured interview instrument: DSM-II = Diagnostic and Statistical Manual of Mental Disorders, 2nd ed. (American Psychiatric
Association 1968); DSM-III = Diagnostic and Statistical Manual of Mental Disorders, 3rd ed. (American Psychiatric Association 1980); ICD-9 = International
Classification of Diseases, 9th version (World Health Organization 1979); NHSI = New Haven Schizophrenia Index (Astrachan et al. 1972); RDC =
Research Diagnostic Criteria (Spitzer et al. 1978)/DIS = Diagnostic Interview Schedule (Robins et al. 1981); SADS = Schedule for Affective Disorders and
Schizophrenia (Endicott and Spitzer 1978).
2
Number of subjects per diagnosis (sex of subject): A = affective disorder; C = nonpsychiatric controls; D = depression; M = mania; MD = manic-
depression; MJD = major depression; N = neurosis; O = other psychiatric disorders; OBS = organic brain syndrome; OC = obsessive-compulsive
disorder; P = personality disorder; PTS = posttraumatic stress disorder; S = schizophrenia; SA = schizoaffective disorder; SP = schizophreniform
disorder. (F) = female; (M) = male.
3
Patient setting where substance abuse was assessed: Al = acute inpatient unit; CI = chronic inpatient unit (over 3 months); ER = emergency room;
I = inpatient ward; O = outpatient setting; VA = Veterans Administration hospital.
4
Definitions of substance abuse/assessment instruments: A = abuse; D = dependence; U = use/DIS = Diagnostic Interview Schedule (Robins et al.
1981); CPS = NIMH Community Support Program Evaluation (McCarrick et al. 1985); MHDS = Manitoba Health and Drinking Survey (Murray 1978).
5
Duration of substance use, abuse, or dependence (e.g., current, past 6 months, lifetime).
6
The percentage of patients in each diagnostic group who abused a particular class of substance is given. ALC = alcohol; STI = stimulants (ampheta-
mines [A], cocaine [C], and related compounds); SED = sedatives (anxiolytics, barbiturates [B], hypnotics, tranquilizers [T]); CAN = cannabis (hashish
[H], marijuana [M], THC); HAL = hallucinogens (LSD [L], MDA, mescaline [M], phencyclidine, psilocybin); NAR = narcotics (codeine, heroin,
morphine, opium); USP = unspecified.
7
Combined sample of single admissions (n = 75), patients with 6 or more admissions (n = 77), and randomly selected patients (n = 85) over a 10- year
period.
8
Males constituted 94% of the total sample.
9
Only the State hospital population is reported here. The specific number of schizophrenic patients assessed in the study of psychiatric and general
hospital admissions was not described in the original article.
10
Abuse determined by urine analysis using thin layer chromatography.
II
Fourteen schizophrenic patients who had never used drugs were compared with 26 who had drug use before their illness developed. Three
schizophrenic patients who used drugs only after the illness developed were dropped, as were 3 patients for whom it could not be determined whether
drug abuse preceded onset of the illness (sex of dropped subjects was not specified).
12
Abuse was defined as amphetamine-induced "subtle or dramatic exacerbations in the course of the illness" (Hansell & Willis 1977, p. 1085).
13
Current cannabis abuse assessed by thin layer silica gel chromatography of urine samples performed on patients returning from day passes.
14
Recent cannabis abuse was assessed by both interview and a cannabinoid assay performed on urine samples.
15
All patients also met operationalized criteria for postpsychotic depression (Siris et al. 1981).
16
Ratings of alcohol abuse and dependence were not mutually exclusive.
17
Alcohol abuse was rated as "moderate" drinking; dependence was rated as "severe" or "extremely severe" drinking. Abuse and dependence were rated
mutually exclusive of each other.
18
Use of "street drugs" rated.
19
S included S, SA, SP; MD included MD, MJD, hypomania.
VOL. 16, NO. 1, 1990 41

Percentage of subjects with substance abuse6

Stimulant Sedative Cannabis Hallucinogen Narcotics Unspecified

S:36.8
(Abuse)
A:52.9
(Abuse)
0:10.0
(Abuse)
S:36.8
(Dependence)
A:29.4
(Dependence)
0:50.0
(Dependence)

20
Physicians' judgments of whether substance abuse played a "major factor" or "minor factor" in precipitating hospitalization were
combined.
21
Abuse of opiates, amphetamines, barbiturates, and methadone determined by thin layer chromotography, gas chromatographic
analysis, and radioimmunoassay of urine samples.
22
Examined use of amphetamines, cocaine, cannabis, and LSD using scale developed by Bowers (1977).
23
Abuse and dependence were mutually exclusive of each other.

Several different strategies may be
used to assess substance abuse, in-
cluding interviews with patients,
significant others, and laboratory
testing for pharmacological sub-
stances. Few standardized instru-
ments exist for obtaining informa-
tion about substance abuse from
patients or significant others (e.g.,
see Bernadt and Murray 1986;
Ananth et al. 1989; Barbee et al.
1989; Drake et al. 1989). Both
psychiatric patients and nonpatients
often deny substance abuse even
when there is clear evidence to the
contrary (Rockwell and Ostwald
1968; Magliozzi et al. 1983; Aiken
1986). At the same time, significant
others are often not privy to infor-
mation about patients' substance
abuse. Routine assays performed on
urine and blood samples have
limitations on the substances and
quantities they detect, and they can-
not determine the pattern and fre-
quency of abuse. Only a few of the
studies reviewed used urine tests,
and these were aimed mainly at
determining the prevalence of cur-
rent substance abuse (Rockwell and
Ostwald 1968; Hall et al. 1977;
Magliozzi et al. 1983). Thus, multi-
ple sources of information are likely
to yield the most reliable estimate of
the prevalence of substance abuse,
particularly if treatment providers
with a therapeutic relationship with
the patient are included.
Duration of Abuse. Substance abuse
must be assessed over a specific
period of time, the duration of
which will influence the estimate
of prevalence. As can be seen in
table 1, estimates are usually higher
for lifetime abuse than for the past
6 months or current abuse (e.g.,
Negrete et al. 1986). Whether recent
or past substance abuse is being
assessed may also affect the reliabil-
ity of the judgments. If one has an
accurate informant (patient or
significant other), information on re-
cent abuse may be more reliable,
since the passage of time can easily
distort memories. On the other
hand, patients may be more moti-
vated to deny recent substance
abuse, for fear of potential negative
consequences (e.g., loss of housing
at home or supervised living ar-
rangement, disappointment or anger
from a mental health worker,
blockage of attempts to enroll in
rehabilitation programs). Awareness
of these potential biases and the
need to obtain information from a
variety of sources will maximize
42
reliable assessment of substance
abuse.
Classes of Substance Abuse.
Previous research has suggested that
schizophrenic patients who are
substance abusers tend to abuse a
wide variety of different drugs
(Blumenfield and Glickman 1967;
Breakey et al. 1974). Drake et al.
(1989) have commented on the dif-
ficulty of accurately assessing the
abuse of different classes of drugs in
this population. There are important
reasons for attempting to assess the
abuse of different classes of sub-
stances. It has been argued that
schizophrenic patients tend to "self-
medicate7' their symptoms by prefer-
entially abusing certain types of
substances, particularly stimulants
and hallucinogens (Pope 1979;
Khantzian 1985; Schneier and Siris
1987; Siris et al. 1988). Only the
assessment of specific classes of
substances can address this
hypothesis.
Of the studies reviewed, 6 examin-
ed unspecified classes of substance
abuse, 5 examined only alcohol
abuse, and 11 examined at least one
specific drug class other than
alcohol. However, only 4 of the 11
studies assessed abuse of more than
one class of drug (Breakey et al.
1974; McLellan and Druley 1977;
Siris et al. 1988; Barbee et al. 1989).
Despite the use of different meth-
odologies, the majority of these
studies indicate that schizophrenic
patients are more prone to abuse
stimulants than other drugs, par-
ticularly when compared with affec-
tive disorder patients. Hallucinogen
abuse may also be higher among
schizophrenic patients, although
fewer studies have assessed this
question and the trend is less clear
(Breakey et al. 1974; McLellan and
Druley 1977; Siris et al. 1988; Barbee
et al. 1989). There is some question
as to whether schizophrenic patients
are more likely to abuse cannabis
than others, but only three studies
examined cannabis abuse in non-
schizophrenic patients (Cohen and
Klein 1970; Breakey et al. 1974;
Magliozzi et al. 1983). The evidence
suggests that schizophrenic patients
are not more likely to abuse alcohol,
sedatives, or narcotics than other pa-
tients or controls.
To examine further the prevalence
of alcohol and drug abuse in
schizophrenia-spectrum disorders,
we assessed the history of substance
abuse in a large sample of rigorously
diagnosed schizophrenic, schizo-
affective, and schizophreniform
disorder patients. The present study
focused on the relations between the
abuse of specific classes of sub-
stances and demographic variables,
the history of the illness, and cur-
rent domains of functioning (symp-
toms and social adjustment).
Methods
The subjects were 149 patients with
diagnoses of schizophrenia (101),
schizoaffective (42), or schizophreni-
form disorder (6), ages 18 to 56, who
were consecutively admitted to
Eastern Pennsylvania Psychiatric In-
stitute for treatment of an acute
exacerbation and who consented to
participate in any of the ongoing
research projects. Fifty-eight subjects
had participated in biological or
pharmacological studies while acute-
ly ill, and had been assessed with
the Schedule for Affective Disorders
and Schizophrenia (SADS) interview
(Endicott and Spitzer 1978); 91 pa-
tients had entered other studies (of
psychosocial variables or of out-
patient treatment), and had been
assessed by the SADS or the Struc-
tured Clinical Interview for DSM-III
(SCID; Spitzer and Williams 1985).
SCHIZOPHRENIA BULLETIN
In the present study, diagnoses were
based on DSM-III-R criteria
generated by the SCID interview or
by review of the SADS ratings and
recorded history.
The majority of patients were
voluntary admissions (76 percent),
single (94 percent), white (54 per-
cent), male (64 percent), and were
living with relatives (75 percent).
The mean age was 30.3 (SD = 8.9)
with a first hospitalization at 22.3
(SD = 6.0) years old. Patients had a
mean of 3.8 (SD = 3.4) prior hospital-
izations and remained in the
hospital for 32.3 (SD = 16.2) days
for the current admission.
DSM-III-R criteria were applied
retrospectively to information
gathered by staff from patient and
family interviews, as well as charts
and past records. Since all patients
were assessed using structured in-
terviews (SADS or SCID), informa-
tion about drug abuse was obtained
directly from patients and in most
cases family members as well.
Abuse of alcohol, stimulants,
sedatives, cannabis, hallucinogens,
and narcotics was assessed (see
table 1, footnote 6, for specific drugs
included in each category). For each
drug category, substance abuse was
rated as either recently present
(patient abused drug within the past
6 months), ever present (patient
abused drug sometime during his/
her life), absent, or unknown. Pa-
tients for whom a history of recent
or lifetime abuse could not be deter-
mined for a particular drug were
dropped from analyses involving
that drug, resulting in some varia-
tion in sample sizes between analy-
ses of different drug classes.
Symptomatology was assessed
1 to 3 weeks after admission using
the Brief Psychiatric Rating Scale
(BPRS; Overall and Gorham 1962)
and the Scale for the Assessment
of Negative Symptoms (SANS;
VOL. 16, NO. 1, 1990
Andreasen 1982). Analyses were
performed using the BPRS subscales
(Anxiety-Depression, Anergia, Ac-
tivation, Hostility, and Thought
Disorder) and the summary scores
for the SANS (Blunted Affect,
Alogia, Apathy, Asociality, and At-
tention). Social adjustment was
rated based on the Social Adjust-
ment Scale-II (SAS-II; Schooler et al.
1979). The following SAS-II sub-
scales were examined in the
analyses: Work, Household, Social-
Leisure, Instrumental Role Function-
ing, and General Adjustment. All
BPRS, SANS, and SAS-II inter-
viewers were trained to interrater
reliabilities of at least Pearson
r = 0.80 on all subscales before they
interviewed study patients. Clinical
ratings of symptoms and social
adjustment were performed by inter-
viewers without knowledge of pa-
tients' history of substance abuse.
Results
To protect against "alpha inflation"
due to multiple statistical tests,
Bonferroni bounds (Kleinbaum et al.
1988) were calculated for an overall a
< 0.05 on the basis of the total
number of statistical tests conducted
(74 tests). Effects that were signifi-
cant at or beyond p < 0.0007 are
significant at the Bonferroni ad-
justed critical p < 0.05 and are
reported here as statistically signifi-
cant. Statistical tests significant at
the unadjusted p < 0.05 level are
reported as trends, along with the
unadjusted probability level.
Pattern of Substance Abuse. Patients
were most likely to have abused
alcohol (47 percent), followed by
cannabis (42 percent), stimulants (25
percent), and hallucinogens (18 per-
cent). Relatively few patients had
abused sedatives (7 percent) or nar-
43
cotics (4 percent). Schizophrenic drug classes, with stimulant, can-
patients were more likely than nabis, and hallucinogen abuse being
schizoaffective disorder patients to most strongly related to each other
have abused amphetamines but did and to other drugs.
not differ in cocaine abuse (figure 1). Statistical analyses were first con-
To determine the degree of co- ducted to examine the relationship
variation in abuse between the dif- between substance abuse and cate-
ferent substances, correlations gorical variables (diagnosis, demo-
(0 coefficients) were computed be- graphics), then continuous variables
tween the six drug categories for (demographics, chronicity), and
both recent (within past 6 months) finally symptom and social adjust-
and lifetime abuse (table 2). Inspec- ment measures.
tion of table 2 indicates moderate in- Categorical and Continuous
tercorrelations among the different Variables. The number and percent-
Figure 1. Percentage of schizophrenic and schizoaffective patients
with a history of amphetamine abuse (left) and cocaine abuse
(right)
Percent of Sample
Abusing Substance
30-
20-
10-
Q
ro
in CM ro IT)
C\J
Amphetamine Cocaine Amphetamine Cocaine
Schizophrenics Schizoaffectives
More schizophrenic patients abused amphetamines as compared to schizoaffective patients (x2 =
5.8, n = 136, df = 1, p < 0.02), but the two groups did not differ in cocaine abuse (x2 < 1, n = 135,
df = 1, NS).
44
Table 2. Intercorrelations among 6 drug categories for the total sample, for lifetime and recent usage
Alcohol
Alcohol
Stimulants 0.171
Sedatives 0.13
Cannabis 0.392
Hallucinogens 0.221
Narcotics 0.211
Note.Tabled are 0 coefficients between drug categories. Coefficients below the diagonal are for lifetime use, and coefficients above the diagonal are
for recent use.
l
p < 0.05, unadjusted.
2
p < 0.05, Bonferroni-adjusted.
ages of patients in each diagnostic
and demographic category who had
recently or ever abused each class of
substances are displayed in table 3.
Linear stepwise discriminant
analyses were conducted to examine
the relationship between a history
of substance abuse and the follow-
ing categorical variables: diagnosis,
gender, race, marital status, legal
status, and preadmission living
arrangement. To limit the number
of analyses performed, recent
substance abuse was not examined.
Six separate analyses were perform-
ed, one for each categorical variable
as the dependent variable (e.g.,
diagnosis, gender), with the same
set of drug classes (e.g., alcohol,
sedatives) in each analysis as the in-
dependent variables. These analyses
yielded information about which
specific types of substance abuse
were most related to each
categorical variable, and are sum-
marized in table 4.
Linear stepwise multiple regres-
sions, similar to the multiple
discriminant analyses described
above, were performed to examine
the relationship between history of
substance abuse (i.e., lifetime
abuse) and the following continuous
variables: age, socioeconomic status
Stimulants Sedatives Cannabis
0.02 0.03 0.292
0.261 0.352
0.292 0.09
0.542 0.221
0.392 0.432 0.482
0.352 0.261 0.231
(level of school factor; Hollingshead
and Redlich 1958), number of prior
hospitalizations, age at first
hospitalization, and length of cur-
rent hospital stay. Five separate
analyses were conducted, one for
each continuous variable as the
dependent variable (e.g., age,
socioeconomic status), with the set
of six drug classes as the indepen-
dent variables in each analysis.
These analyses are summarized in
table 5.
Demographic Characteristics.
Gender, age, race, and socio-
economic status were all related to
specific types of substance abuse.
Males abused each class of drugs
more than females did, particularly
alcohol and cannabis. Young pa-
tients were more prone to abuse
stimulants (both cocaine and am-
phetamines) than other drugs. There
was an interaction between race and
different types of substance abuse.
White patients were more likely to
have abused alcohol or sedatives
and less likely to have abused can-
nabis than black patients were.
Within the sedative abusers, white
patients were more likely to have
abused barbiturates than blacks
were (x2 = 7.3, n = 138, df = 1,
SCHIZOPHRENIA BULLETIN
Hallucinogens Narcotics
0.08 0.12
0.502 -0.03
0.201 -0.01
0.271 0.16
-0.02
0.322
p < 0.007, unadjusted), but not
anxiolytics (x2 = 3.5, n = 138,
df = 1, NS). Finally, patients with a
lower socioeconomic status were
more likely to have abused
cannabis.
Diagnosis and History of Illness.
Schizophrenic patients showed a
greater tendency to have a history
of stimulant abuse than schizoaffec-
tive patients did. Stimulant abuse
was also associated with an earlier
age at first hospitalization, but not a
greater number of hospitalizations.
Cocaine and amphetamine abuse
were not differentially related to age
of first hospitalization, as determin-
ed by analysis of variance (ANOVA).
Against expectations, a history of
cannabis abuse was related to fewer
hospitalizations.
Symptoms and Social Adjustment.
The relationship of substance abuse
to symptomatology and social ad-
justment was examined with linear
stepwise multiple discriminant
analyses. Three separate analyses
were conducted for the abuse history
of each drug class, one for each
rating scale (BPRS, SANS, and
SAS-II). For each analysis, the
subscales on the measure (e.g., the
 CD
z
p
Table 3. Number and corresponding percentage of patients who recently used (RU) or ever used (EU) each of 6
drug categories, for diagnostic and demographic variables -"*
VOL. 16, NO. 1, 1990

Alcohol Stimulants Sedatives Cannabis Hallucinogens Narcotics

RU EU RU EU RU EU RU EU RU EU RU EU
Groups n (%) n (o/o) n (o/o) n (o/o) n (o/o) n (o/o) n (o/o) n (o/o) n (o/o) n (o/o) n (o/o) n (o/o)

Total sample 46 (33) 66 (47) 20 (14) 35 (25) 6 (4) 10 ( 7) 31 (22) 58 (42) 6 (4) 25 (18) 1 ( 1) 6 ( 4)

Schizophrenic 34 (35) 48 (50) 15 (16) 28 (29)1 4 (3) 6 ( 6) 23 (23) 40 (43) 6 ( 6) 19 (20) 1 ( 1) 6 ( 5)

Schizophreniform 2 (40) 3 (60) 2 (40) 2 (40) 0 (0) 0 (0) 3 (60) 4 (80) 0 (0) 0 (0) 0 (0) 0 ( 0)
Schizoaffective 10 (26) 15 (38) 3 (7) 5 (12)1 2 ( 5) 4 (10) 5 (13) 14 (35) 0 (0) 6 (15) 0 (0) 0 ( 0)

Male 37 (41) 52 (57)2 19 (21) 30 (33) 5 (4) 7 (8) 28 (30) 50 (55)2 5 (5) 20 (22) 1 ( 1) 4 (3)
Female 9 (18) 14 (29)2 1 (2) 5 (10) 1 (2) 3 ( 6) 3 ( 6) 8 (17)2 1 ( 2) 5 (10) 0 (0) 2 (4)

White 30 (38) 44 (56)3 13 (17) 20 (27) 6 (7) 10 (13)3 5 (19) 29 (39)3 6 ( 8) 19 (25) 0 (0) 5 ( 5)
Black 16 (25) 22 (34)3 7 (11) 15 (23) 0 (0) 0 (0)3 16 (26) 29 (47)3 0 (0) 6 (9) 1 (2) 1 (2)

Single 45 (34) 65 (49) 19 (14) 34 (26) 6 ( 4) 10 ( 8) 30 (22) 56 (43) 5 (4) 24 (18) 1 (1) 6 (4)
Married 1 (14) 1 (14) 1 (13) 1 (13) 0 (0) 0 (0) 1 (13) 2 (25) 1 (13) 1 (13) 0 (0) 0 (0)

Voluntary 37 (35) 53 (50) 16 (15) 26 (24) 6 (5) 9 ( 8) 22 (20) 45 (42) 4 ( 4) 18 (17) 1 ( 1) 4 ( 3)

Involuntary 9 (26) 13 (38) 4 (13) 9 (28) 0 (0) 1 (3) 9 (28) 13 (41) 2 ( 6) 7 (21) 0 (0) 2 (6)

Family home 39 (34) 56 (49) 19 (16) 31 (27) 6 (4) 8 ( 7) 27 (24) 50 (44) 5 (4) 22 (19) 1 ( 1) 6 (4)
Boarding home 7 (28) 10 (40) 1 (4) 4 (15) 0 (0) 2 ( 8) 3 (12) 8 (31) 1 (4) 3 (12) 0 (0) 0 (0)

Note.Recent use (RU) = within the past 6 months; Ever used (EU) = any prior abuse (both RU and EU scored as Yes [1] or No [0]). Tabled are the number (n) of each
demographic class that were Yes for RU or EU, and the corresponding percentage (%) of the diagnostic or demographic class that n reflects. See Methods for definitions of
drug categories, and of marital, legal, and living status.
1
Trend for schizophrenic patients to abuse more stimulants than schizoaffective patients did in linear stepwise discriminant analysis (LSDA; p < 0.05, unadjusted).
2
Significant effects for males to abuse more alcohol and cannabis than females in LSDA (p < 0.05).
3
Significant effects for whites to abuse more alcohol and sedatives and less cannabis than blacks in LSDA (p < 0.05).
45
 46 SCHIZOPHRENIA BULLETIN

Table 4. Discriminating diagnostic group and dichotomous demographic classes using 6 drug
categories (n = 124)
Druqs
Dependent
variable Classes Code ALC STI SED CAN HAL NAR

Diagnosis1 Schizophrenia 1 4.44 0.03

Schizoaffective 0
Sex Male 1 13.95 0.18 + +
Female 0
Race2 White 1 7.55 0.20 + + -
Black 0
Marital status Married 1
Single 0
Legal status Voluntary 1
Involuntary 0
Living arrangement3 With family 1
Boarding house 0
Note.ALC = alcohol; STI = stimulants; SED = sedatives; CAN = cannabis; HAL = hallucinogens; NAR = narcotics (see text for definition of drug
categories). Tabled are the signs for ft coefficients for drug categories that emerged as statistically significant components of the linear stepwise discrimi-
nant model (LSDM). A dummy-code (1 or 0) was assigned to each class during statistical analysis. F values are based on LSDMs, with drug categories
treated as independent variables. Analyses by stepwise logistic regression yielded completely parallel conclusions. Dashes indicate that no model fit with
p < 0.05 by either criterion for a Type I error.
1
The 6 schizophreniform patients were dropped from this analysis due to the small number.
2
The 2 Oriental patients were dropped from this analysis.
3
The 3 patients who lived on the streets were dropped from this analysis, as were the 14 patients who lived alone and 3 who lived with friends.
*p < 0.05, unadjusted.
5
p< 0.05, Bonferroni-adjusted.

Table 5. Prediction of continuous demographic and chronicity variables using 6 drug categories
(n = 124)
Druqs
Dependent variable ALC STI SED CAN HAL NAR

Age (years) 12.22 o.O9

1
Socioeconomic status 8.43 0.06
Number of prior hospitalizations 5.13 o.O4
Age (years) at first hospitalization 7.73 0.06
Length of stay (months)
Note. ALC = alcohol; STI = stimulants; SED = sedatives; CAN = cannabis; HAL = hallucinogens; NAR = narcotics (see text for definition of drug
categories). Tabled are the signs for ft coefficients for drug categories that emerged as statistically significant components of the linear stepwise regres-
sion model (LSRM). The F values are based on LSRMs, with drug categories treated as independent variables. Dashes indicate that no model fit with p <
0.05 by either criterion for a Type I error.
1
Based on Hollingshead-Redlich Scale (1958), in which high numbers reflect lower socioeconomic status.
2
p< 0.05, Bonferroni-adjusted.
3
p< 0.05, unadjusted.
VOL 16, NO. 1, 1990
BPRS) were used to examine
whether the patient had a positive
history of abuse for the drug class.
Discriminant analyses were also
performed on recent abuse for
the three classes of drugs that had
been recently abused by 10 or more
patients: alcohol, cannabis, and
stimulants (see table 3).
No effects were found for
stimulant or narcotic abuse on any
clinical variables. The abuse of some
drugs was related to lower levels of
symptomatology. A history of can-
nabis abuse was related to signifi-
cantly lower scores on the Activa-
tion subscale of the BPRS (F = 13.1;
df = 1, 82; p < 0.05), and there was
a trend for lower Asociality scores
on the SANS (F = 6.4; df = 1, 112;
p < 0.05, unadjusted). There was
also a trend for patients who had
abused hallucinogens to have lower
Anxiety-Depression scores on the
BPRS (F = 7.2; df = 1.83; p < 0.05,
unadjusted).
However, there were trends for
patients who had recently (F = 7.4)
or ever (F = 4.0) abused alcohol to
have worse Blunted Affect as
measured by the SANS (df = 1, 114,
p < 0.05, unadjusted). There were
also trends for patients who had
ever abused alcohol to have better
Work Adjustment on the SAS-II
(F = 7.2; df = 1, 77; p < 0.05, un-
adjusted), but for recent alcohol
abusers to have worse General
Adjustment on the same instrument
(F = 6.1; df = 1, 77) p < 0.05, un-
adjusted). Finally, a trend was
present for patients who had
abused sedatives to have higher
Thought Disorder scores on the
BPRS (F = 4.6; df = 1, 82; p < 0.05,
unadjusted).
Discussion
Prevalence and Patterns of
Substance Abuse. The present study
was limited by the fact that retro-
spective ratings of substance abuse
were made, and that abuse was not
assessed in a matched group of
psychiatric patients with diagnoses
other than schizophrenia-spectrum
disorders or nonpatient controls.
The strengths of the study were the
rigorous diagnostic procedures, the
relatively large sample size, the
assessment of multiple classes of
substance abuse, and the use of
statistical analyses that examined
different classes of substance abuse
simultaneously. Thus, caution must
be exercised when comparing
prevalence rates reported here with
those for other populations.
A rough comparison can be made
between the prevalence of substance
abuse in this group of patients and
that of the general population of
persons with similar demographic
characteristics living in the North-
east United States in 1985, as
established in the National House-
hold Survey on Drug Abuse (Na-
tional Institute on Drug Abuse
1987b). Schizophrenic patients had
higher rates of hallucinogen and
stimulant abuse, particularly am-
phetamine abuse, but lower rates of
cannabis, sedative, alcohol, and nar-
cotic abuse. For example, 25 percent
of the schizophrenic patients had
abused amphetamines and 20 per-
cent had abused hallucinogens,
compared to only 15 percent of the
general population for each drug. In
contrast, only 6 percent of the
schizophrenic patients had abused
sedatives and 5 percent had abused
narcotics, compared to 11 percent
and 9 percent of the general popula-
tion, respectively. It should be noted
that the abuse rates cited in the
National Household Survey are bas-
ed on self-reports of psychoactive
substance use, and hence are prob-
ably overestimates of the prevalence
47
of substance abuse disorders in the
general population, as defined by
DSM-III-R.
These findings are consistent with
a recent report of patients seeking
assistance for problems with
substance abuse (Ross et al. 1988), in
which 22 DSM-III schizophrenic pa-
tients had a higher prevalence of
amphetamine and hallucinogen
abuse than other drugs compared
with nonschizophrenic patients.
These results are also in line with
the review by Schneier and Siris
(1987) of substance abuse in schizo-
phrenia and with the studies
reviewed here. In addition to the
limitations cited above, however, it is
also possible that stimulant abuse
was overestimated in the present
study, since amphetamine abuse
may be more prevalent in Philadel-
phia than other urban areas in the
Northeast United States (National
Institute on Drug Abuse 1987a). Also,
most of the data collected here
predate the recent increase in co-
caine abuse in the general popula-
tion. It is unknown whether
schizophrenic patients exhibit a
greater preference for cocaine than
the general population does when it
is more readily available.
The moderate correlations for re-
cent and lifetime abuse of the six
drug classes (table 2) indicate that
patients who abused drugs used a
wide range of different substances.
Schizophrenic patients have been
found to abuse a greater variety of
drugs than other patients and con-
trols (Blumenfield and Glickman
1967; Breakey et al. 1974), although
the quantity of drugs abused tends
to be lower (Cohen and Klein 1970;
Crowley et al. 1974; Ritzier et al.
1977). Abuse of cannabis, hallucino-
gens, and stimulants was more
highly correlated with each other
than with other drugs. However,
abuse of specific drugs did not
48
appear to be determined by the
similarity of their effects: recent and
lifetime histories of alcohol and
sedative abuse were not correlated
with each other, whereas recent and
lifetime histories of abuse of stimu-
lants and sedatives were. These data
differ from those reported by
McLellan et al. (1985), who found
that psychiatric inpatients with a
history of substance abuse tended to
use combinations of drugs with
similar, rather than opposing effects.
However, their patients were heavy
substance abusers with use occur-
ring at least three times weekly, and
schizophrenic patients constituted
only half of the sample.
Demographic Correlates. As in
research conducted with nonpsychi-
atric patients, demographic char-
acteristics were strongly correlated
with substance abuse. Gender, age,
and race were significantly related to
history of substance abuse at the
Bonferroni-corrected level of p <
0.0007. Socioeconomic status was
marginally related to drug abuse
(i.e., at the uncorrected level of p <
0.05), whereas marital status, legal
status at admission (voluntary/in-
voluntary), and living arrangement
were not.
Males were more prone to abuse
all substances than females were
(table 3), with alcohol and cannabis
abuse being significantly higher
(table 4) in the multiple discriminant
analysis. This difference is in line
with an epidemiological study of
psychiatric disorder in three dif-
ferent urban areas (Myers et al.
1984), in which males had twice the
rate of drug abuse or dependence,
and more than four times as much
alcohol abuse or dependence as
females. Many studies with
psychiatric patients have also
reported more substance abuse in
males than females (Hensala et al.
1967; Robinson and Wolkind 1970;
Chopra and Smith 1974; Hall et al.
1977; Bowers and Swigar 1983;
Walker et al. 1985; Negrete et al.
1986; Solomon 1986; Drake et al.
1989). A few reports have found no
gender difference in substance
abuse for psychiatric patients
(Fischer et al. 1975; Westermeyer
and Walzer 1975; Safer 1987; Barbee
et al. 1989). In the present study,
only alcohol and cannabis abuse
were independently related to
gender. Thus, both male and female
patients who had abused hallucino-
gens, stimulants, sedatives, and nar-
cotics also tended to have a positive
history for alcohol abuse, cannabis
abuse, or both.
Age was related to substance
abuse, with younger patients abus-
ing significantly more stimulants
(table 5). Young psychiatric patients
have been found to abuse more
drugs in most studies (Hensala et al.
1967; Robinson and Wolkind 1970;
Breakey et al. 1974; Crowley et al.
1974; Fischer et al. 1975; McLellan et
al. 1978; Alterman et al. 1981, 1982;
O'Farrell et al. 1983; Richard et al.
1985; Negrete et al. 1986; Siris et al.
1988; Drake et al. 1989). While two
reports have documented high rates
of stimulant abuse among psychi-
atric inpatients (Robinson and
Wolkind 1970) and schizophrenic
admissions to an inpatient unit
(Richard et al. 1985), these studies
did not assess the prevalence of
other types of substance abuse.
Since only a history of stimulant
abuse was related to age in our
study, a general propensity toward
substance abuse among younger pa-
tients would not explain this effect.
The increased availability of stimu-
lants over the past decade (Miller
et al. 1983) is one possible explana-
tion for greater stimulant abuse
among younger patients.
SCHIZOPHRENIA BULLETIN
Race was a significant determinant
of substance abuse history. In this
sample, white patients were more
likely to have abused alcohol or
hallucinogens, and less likely to
have abused cannabis than black pa-
tients. However, previous studies
that investigated racial differences in
substance abuse among schizo-
phrenic patients are not uniform in
their results. Pokorny (1965) found
more alcoholism and Hensala et al.
(1967) found more LSD use in white
than black psychiatric admissions.
On the other hand, Alterman et al.
(1981, 1982) reported greater covert
drug abuse among black than white
inpatients. While other studies have
failed to find racial differences in
drug abuse among psychiatric pa-
tients (Fischer et al. 1975;
Westermeyer and Walzer 1975; Hall
et al. 1977; Barbee et al. 1989), those
studies did not examine substance
abuse of different types of drugs
among patients with schizophrenia.
The interaction between race and
drug type found in the current sam-
ple underscores the importance of
examining racial differences in the
patterns of abuse of different drugs;
no difference between the races
would have been found if overall
substance abuse had been studied.
The strong relationship between
demographic factors and substance
abuse suggests that the environment
is probably a critical determinant of
who abuses which drugs. The
prevalence of substance abuse in an
individual's social environment, in-
cluding their ethnic and peer group,
influences the availability of dif-
ferent drugs, and hence the likeli-
hood that they will be abused.
These results indicate that the same
environmental effects on drug abuse
reported in the nonpsychiatric
population also influence the pat-
tern of substance abuse among
VOL. 16, NO. 1, 1990
schizophrenic patients (Spieger and
Harford 1987).
Clinical Correlates. Patients with a
history of cannabis abuse had
significantly lower Activation scores
on the BPRS (tension, mannerisms,
and excitement) and tended to have
lower Asociality scores on the SANS
and fewer previous hospitalizations.
While others have reported that can-
nabis abuse can worsen the symp-
toms of schizophrenia (Bernhardson
and Gunne 1972; Treffert 1978;
Knudsen and Vilmar 1984), recent
cannabis abuse (within the past
6 months) was not related to symp-
toms in this study. The milder
clinical severity of patients who had
abused cannabis may reflect the in-
teraction between social competence
and environmental influences. Can-
nabis is often used in a social peer
group, and a "culture" exists around
its consumption in this society
(Becker 1953). The lower symptoms
of patients who had a history of can-
nabis abuse may reflect self-selection
whereby persons who were more
socially competent were more prone
to abuse due to their higher ex-
posure to the drug through social
contacts. Cohen and Klein (1970)
noted that most psychotic patients
lacked the social skills to sustain
heavy drug use, and others have
reported that psychiatric patients
who are alcohol abstainers are more
chronic and have a worse clinical
outcome than those who imbibe
(Ritzier et al. 1977; OTarrell et al.
1983).
Several other trends in the clinical
data deserve brief mention. Caution
should be exercised in interpreting
these results, since they did not
meet the stringent Bonferroni-
adjusted criteria for statistical
significance, although they were
significant at the conventional
p < 0.05 level. A history of stimulant
abuse was found in patients with an
earlier age of first hospitalization.
Some studies have reported an
earlier age of onset of the illness for
drug-abusing schizophrenic patients
(Breakey et al. 1974; Tsuang et al.
1982; Alterman et al. 1984) or psy-
chiatric patients (Westermeyer and
Walzer 1975), but others have not
(Roy 1981; Hays and Aidroos 1986;
Safer 1987). No research, however,
has examined the importance of
stimulant abuse compared with
abuse of other types of drugs to the
age of onset. These data suggest that
stimulant use may precipitate an
earlier onset of illness, although we
did not have information on
whether abuse actually preceded the
illness.
Schizophrenic patients were more
likely to have abused stimulants,
particularly amphetamines, than
were schizoaffective patients,
although the two groups did not dif-
fer in cocaine abuse. Different
mechanisms have been hypothesiz-
ed to underlie the greater abuse of
stimulants by schizophrenic pa-
tients, such as self-medication (e.g.,
Pope 1979; Schneier and Siris 1987)
or a stimulant-induced schizo-
phreniform disorder (McLellan et al.
1979; Bowers 1987). While the design
of the current study does not permit
a test of these theories, the data sug-
gest that schizoaffective patients
may not reflect the increased rate of
stimulant abuse seen in schizo-
phrenic patients.
A final noteworthy trend is the
worse scores of Blunted Affect
(SANS) of patients who had recently
or ever abused alcohol. Many
schizophrenic patients who abuse
alcohol emphatically state that it
ameliorates their symptoms (Alpert
49
and Silvers 1970; Hansell and Willis
1977). Whether blunted affect during
a drug-free state reflects the in-
fluence of chronic alcohol abuse or
persistent symptoms that patients
attempt to self-medicate cannot be
determined. Prospective longi-
tudinal research will be necessary to
address this question.
Conclusions
Substance abuse in schizophrenia is
a significant problem that has im-
portant theoretical and clinical im-
plications for understanding and
managing the course of the illness.
Thus far, most of the research on the
prevalence of substance abuse in
schizophrenia has suffered from
methodological shortcomings that
impede comparisons between
studies and limit the conclusions
that can be drawn. Nevertheless,
there is a suggestion from findings
in the literature (table 1) and data
presented here that schizophrenic
patients are more likely to have a
history of abusing stimulants (par-
ticularly amphetamines) and
hallucinogens, but not alcohol or
other drugs, when compared to
other patient groups. Several factors
are associated with an increased
prevalence of substance abuse, in-
cluding gender, age, race, and
socioeconomic status.
For future epidemiological studies
in this area to be comparable, basic
methodological standards need to be
considered in the design of research
(see Appendix 1). Longitudinal
studies with multiple assessments of
both substance abuse and clinical
variables will be useful in estimating
the prevalence of substance abuse in
schizophrenia and teasing out the
differences between self-medication
and the influence of abuse on the
course of the illness.
 50
References
Aiken, L.S. Retrospective self-
reports by clients differ from original toms in schizophrenia. Archives of
reports: Implications for the evalua-
tion of drug treatment programs.
The International Journal of the Addic-
tions, 21:767-788, 1986.
Alpert, M., and Silvers, K.N.
Perceptual characteristics distin-
guishing auditory hallucinations in
schizophrenia and acute alcoholic
psychosis. American Journal of
Psychiatry, 127:298-302, 1970.
Alterman, A.I.; Ayres, F.R.; and
Williford, WO. Diagnostic validation
of conjoint schizophrenia and alco-
holism. Journal of Clinical Psychiatry, M.; Wolkin, A.; and Rotrosen, J.
45:300-303, 1984.
Alterman, A.I.; Erdlen, D.L.;
LaPorte, D.J.; and Erdlen, F.R.
Effects of illicit drug use in an in-
patient psychiatric population. Ad-
dictive Behaviors, 7:231-242, 1982.
Alterman, A.I.; Erdlen, F.R.; and
Murphy, E. Alcohol abuse in the
psychiatric hospital population.
Addictive Behaviors, 6:69-73, 1981.
American Psychiatric Association.
DSM-II: Diagnostic and Statistical
Manual of Mental Disorders. 2nd ed.
Washington, DC: The Association,
1968.
American Psychiatric Association.
DSM-III: Diagnostic and Statistical
Manual of Mental Disorders. 3rd ed.
Washington, DC: The Association,
1980.
American Psychiatric Association.
DSM-III-R: Diagnostic and Statistical Kehoe, C.E. Alcohol and substance
Manual of Mental Disorders. 3rd ed., abuse among schizophrenic patients
revised. Washington, DC: The Asso-
ciation, 1987.
Ananth, J.; Vandewater, S.; Kamal,
M.; Brodsky, A.; Gamal, R.; and
Miller, M. Mixed diagnosis of sub-
stance abuse in psychiatric patients.
Hospital & Community Psychiatry,
40:297-299, 1989.
Andreasen, N.C. Negative symp-
General Psychiatry, 39:784-788, 1982.
Angermeyer, M.C., and Kuhn, L.
Gender difference in age at onset of
schizophrenia: An overview. Euro-
pean Archives of Psychiatry and
Neurological Sciences, 237:351-364,
1988.
Angrist, B.; Peselow, E.; Rubinstein,
M.; Corwin, J.; and Rotrosen, J. Par-
tial improvement in negative symp-
toms after amphetamine. Psycho-
pharmacology, 78:128-130, 1982.
Angrist, B.; Peselow, E.; Rubinstein,
Amphetamine response and relapse
risk after depot neuroleptic discon-
tinuation. Psychopharmacology,
85:277-283, 1985.
Angrist, B.; Rotrosen, J.; and
Gershon, S. Differential effects of
amphetamine and neuroleptics on
negative vs. positive symptoms of
schizophrenia. Psychopharmacology,
72:17-19, 1980.
Astrachan, B.M.; Harrow, M.; Adler,
D.; Braver, L.; Schwartz, A.;
Schwartz, C ; and Tucker, G.
A checklist for the diagnosis of
schizophrenia. British Journal of
Psychiatry, 121:529-534, 1972.
Atkinson, R.M. Importance of
alcohol and drug abuse in psychi-
atric emergencies. California
Medicine, 118:1-4, 1973.
Barbee, J.G.; Clark, P.D.; Crapan-
zano, M.S.; Heintz, G.C.; and
presenting to an emergency psychi-
atric service. Journal of Nervous and
Mental Disease, 177:400-407, 1989.
Becker, H.S. Becoming a marijuana
user. American Journal of Sociology,
59:235-242, 1953.
SCHIZOPHRENIA BULLETIN
Bell, D.S. Comparison of ampheta-
mine psychosis and schizophrenia.
British Journal of Psychiatry,
1121:701-707, 1965.
Bernadt, M.W., and Murray, R.M.
Psychiatric disorder, drinking and
alcoholism: What are the links?
British Journal of Psychiatry,
148:393-400, 1986.
Bernhardson, G., and Gunne, L.M.
Forty-six cases of psychosis in can-
nabis abusers. International Journal of
the Addictions, 7:9-16, 1972.
Blumenfield, M., and Glickman, L.
Ten months' experience with LSD
users admitted to county psychiatric
receiving hospital. New York State
Journal of Medicine, 67:1849-1853,
1967.
Bowers, M.B., Jr. Acute psychosis in-
duced by psychomimetic drug abuse:
I. Clinical findings. Archives of
General Psychiatry, 27:437-440, 1972.
Bowers, M.B., Jr. Psychosis precipi-
tated by psychotomimetic drugs: A
followup study. Archives of General
Psychiatry, 34:832-835, 1977.
Bowers, M.B., Jr. The role of drugs
in the production of schizophreni-
form psychosis and related dis-
orders. In: Meltzer, H.Y., ed.
Psychopharmacology: The Third
Generation of Progress. New York:
Raven Press, 1987. pp. 819-823.
Bowers, M.B., Jr., and Swigar, M.E.
Vulnerability to psychosis associated
with hallucinogen use. Psychiatry
Research, 9:91-97, 1983.
Breakey, W.R., and Goodell, H.
Thought disorder in mania and
schizophrenia evaluated by Ban-
nister's Grid Test for Schizophrenic
Thought Disorder. British Journal of
Psychiatry, 120:391-395, 1972.
Breakey, W.R.; Goodell, H.; Lorenz,
P C ; and McHugh, PR. Hallucino-
genic drugs as precipitants of schizo-
phrenia. Psychological Medicine,
4:255-261, 1974.
VOL. 16, NO. 1, 1990
Carpenter, M.D.; Mulligan, J.C;
Bader, LA.; and Meinzer, A.E.
Multiple admissions to an urban
psychiatric center. Hospital & Com-
munity Psychiatry, 31:397-400, 1985.
Cesarec, Z., and Nyman, A.K. Dif-
ferential response to amphetamine
in schizophrenia. Ada Psychiatrica
Scandinavica, 71:523-538, 1985.
Chopra, G.S. Marijuana and adverse
psychotic reactions. Bulletin on Nar-
cotics, 23:15-22, 1971.
Chopra, G.S., and Smith, J.W.
Psychotic reactions following can-
nabis use in East Indians. Archives of
General Psychiatry, 30:24-27, 1974.
Clark, L.D.; Hughes, R.; and
Nakashima, E.H. Behavioral effects
of marijuana. Archives of General
Psychiatry, 23:193-198, 1970.
Cohen, M., and Klein, D.E Drug
abuse in a young psychiatric popula-
tion. American Journal of Ortho-
psychiatry, 40:448-455, 1970.
Cohen, S. Suicide following morn-
ing glory seed ingestion. American
Journal of Pyschiatry, 120:1024-1025,
1964.
Cohen, S. A classification of LSD
complications. Psychosomatics,
7:182-186, 1966.
Crowley, T.J.; Chesluk, D.; Dilts, S.;
and Hart, R. Drug and alcohol
abuse among psychiatric admis-
sions. Archives of General Psychiatry,
30:13-20, 1974.
Davidoff, E., and Reifenstein, E.C.,
Jr. The results of eighteen months of
benzedrine sulfate therapy in psy-
chiatry. American Journal of Psychiatry,
95:945-970, 1939.
DeVito, R.A.; Flaherty, L.A.; and
Mozdzierz, G.J. Toward a psycho-
dynamic theory of alcoholism.
Diseases of the Nervous System,
31:43-49, 1970.
Diethelm, O. Some fundamental
considerations in the treatment of
chronic alcoholism. In: Himwich,
H.E., ed. Alcoholism, Basic Aspects
and Treatment. Pub. No. 47.
Washington, DC: American Associa-
tion for the Advancement of
Science/1957 pp. 191-198.
Donovan, D.M., and Marlatt, D.A.,
eds. Assessment of Addictive Behaviors.
New York: Guilford Press, 1988.
Drake, R.E.; Osher, EC; and
Wallach, M.A. Alcohol use and
abuse in schizophrenia: A prospec-
tive community study. Journal of
Nervous and Mental Disease,
177:408-414, 1989.
Endicott, ]., and Spitzer, R.L.
A diagnostic interview: The
Schedule for Affective Disorders and
Schizophrenia. Archives of General
Psychiatry, 35:837-844, 1978.
Feighner, J.P.; Robins, E.; Guze,
S.B.; Woodruff, R.A.; Winoker, G.;
and Munoz, R. Diagnostic criteria
for use in psychiatric research.
Archives of General Psychiatry,
26:57-63, 1972.
Fischer, D.E.; Halikas, J.A.; Baker,
J.W; and Smith, J.B. Frequency and
patterns of drug abuse in psychiatric
patients. Diseases of the Nervous
System, 36:550-553, 1975.
Freed, E.X. Alcoholism and schizo-
phrenia: The search for perspec-
tives. Journal of Studies on Alcohol,
36:853-881, 1975.
Freedman, D.X. On the use and
abuse of LSD. Archives of General
Psychiatry, 18:330-347, 1968.
Galanter, M.; Castaneda, R.; and
Ferman J. Substance abuse among
general psychiatric patients: Place of
presentation, diagnosis and treat-
ment. American Journal of Drug and
Alcohol Abuse, 142:211-235, 1988.
51
Ghodse, A.H. Cannabis psychosis.
British Journal of Addiction,
81:473-478, 1986.
Goldstein, J.M. Gender differences
in the course of schizophrenia.
American Journal of Psychiatry,
145:684-689, 1988.
Hall, R.C.W; Popkin, M.K.;
Beresford, T.P.; and Hall, A.K.
Amphetamine psychosis: Clinical
presentations and differential
diagnosis. Psychiatric Medicine,
6:73-79, 1988.
Hall, R.C.W; Popkin, M.K.; and
DeVaul, R. The effect of unrecog-
nized drug abuse on diagnosis and
therapeutic outcome. American Jour-
nal of Drug and Alcohol Abuse,
4:455-465, 1977.
Hansell, N., and Willis, G.L. Out-
patient treatment of schizophrenia.
American Journal of Psychiatry,
134:1082-1086, 1977.
Hays, P., and Aidroos, N. Alcohol-
ism followed by schizophrenia. Ada
Psychiatrica Scandinavica, 74:187-189,
1986.
Helzer, J.E.; Clayton, P.J.; Pam-
bakian, R.; Reich, T.; Woodruff,
R.A.; and Reveley, M.A. Reliability
of psychiatric diagnosis: II. Archives
of General Psychiatry, 34:136-141,
1977.
Hensala, J.D.; Epstein, L.J.; and
Blackel, K.H. LSD and psychiatric
inpatients. Archives of General
Psychiatry, 16:554-559, 1967.
Hollingshead, A.B., and Redlich,
EC. Social Class and Mental Illness: A
Community Study. New York: John
Wiley & Sons, Inc., 1958.
Hollister, L.E. Health aspects of can-
nabis. Pharmacology Review, 38:1-20,
1986.
Janowsky, D.S., and Davis, J.M.
Methylphenidate, dextroampheta-
52
mine, and levamphetamine. Archives
of General Psychiatry, 33:304-308,
1976.
Janowsky, D.S.; El-Yousef, M.K.;
Davis, J.M.; and Sekerke, H.J.
Provocation of schizophrenic symp-
toms by intravenous administration
of methylphenidate. Archives of
General Psychiatry, 28:185-191, 1973.
Kesselman, M.S.; Solomon, J.;
Beaudett, M.; and Thornton, B.
Alcoholism and schizophrenia. In:
Solomon, J., ed. Alcoholism and
Clinical Psychiatry. New York:
Plenum Press, 1982. pp. 69-80.
Khantzian, E.J. The self-medication
hypothesis of addictive disorders:
Focus on heroin and cocaine depen-
dence. American Journal of Psychiatry,
142:1259-1264, 1985.
Klein, D. Importance of psychiatric
diagnosis in prediction of clinical
drug effects. Archives of General
Psychiatry, 16:118-126, 1967.
Kleinbaum, D.G.; Kupper, L.L.; and
Muller, K. Applied Regression Analysis
and Other Multivariable Methods.
Boston: PWS-Kent Publishing Com-
pany, 1988.
Knudsen, P., and Vilmar, T. Can-
nabis and neuroleptic agents in
schizophrenia. Ada Psychiatrica Scan-
dinavica, 69:162-174, 1984.
Lewis, N., and Piotrowski, Z.A.
Clinical diagnosis of manic-
depressive psychosis. In: Hoch, P.,
and Zubin, J., eds. Depression. New
York: Grune & Stratton, 1954.
pp. 25-38.
Lieberman, J.A.; Kane, J.M.;
Sarantakos, S.; Gadaleta, D.;
Woerner, M.; Alvir, J.; and Ramos-
Lorenzi, J. Prediction of relapse in
schizophrenia. Archives of General
Psychiatry, 44:597-603, 1987.
Magliozzi, J.R.; Kanter, S.L.; Cser-
nansky, J.G.; and Hollister, L.E.
Detection of marijuana use in psy-
chiatric patients by determination of
urinary delta-9-tetrahydrocannabino-
11-oic acid. Journal of Nervous and
Mental Disease, 171:246-249, 1983.
McCarricky A.K.; Manderscheid,
R.W.; and Bertolucci, D.E. Correlates
of acting out behaviors among
young adult chronic patients.
Hospital & Community Psychiatry,
36:848-853, 1985.
McLellan, A.T.; Childress, A.R.; and
Woody, G.E. Drug abuse and psy-
chiatric disorders. In: Alterman,
A.I., ed. Substance Abuse and
Psychopathology. New York: Plenum
Press, 1985. pp. 137-172.
McLellan, AX, and Druley, K.A.
Non-random relation between drugs
of abuse and psychiatric diagnosis.
journal of Psychiatric Research,
13:179-184, 1977.
McLellan, A.T.; Druley, K.A.; and
Carson, J.C. Evaluation of substance
abuse problems in a psychiatric
hospital. Journal of Clinical Psychiatry,
39:425-430, 1978.
McLellan, A.T.; Woody, G.E.; and
O'Brien, C.P. Development of
psychiatric illness in drug abusers:
Possible role of drug preference. New
England Journal of Medicine,
301:1310-1314, 1979.
Miller, J.D.; Cisin, I.H.; Gardner-
Keaton, H.; Harrell, A.V.; Wirtz,
P.W.; Abelson, H.I.; and Fishourne,
P.M. National Survey on Drug Abuse:
Main Findings 1982. DHHS Pub. No.
(ADM) 83-1263. Washington, DC:
National Institute on Drug Abuse,
1983.
Mott, R.H.; Small, I.F.; and Ander-
son, J.M. A comparative study of
hallucinations. Archives of General
Psychiatry, 12:595-601, 1965.
Murray, R.P Manitoba Health and
Drinking Survey: The Social Epidemi-
SCHIZOPHRENIA BULLETIN
ology of Alcohol Use in Eastman
Region. Winnipeg: University of
Winnipeg, 1978.
Myers, J.K.; Weissman, M.M.;
Tischler, G.L.; Holzer, C.E. Ill; Leaf,
P.J.; Orvaschel, H.; Anthony, J.C;
Boyd, J.H.; Burke, J.D., Jr.; Kramer,
M.; and Stoltzman, R. Six-month
prevalence of psychiatric disorders
in three communities. Archives of
General Psychiatry, 41:959-967, 1984.
National Institute on Drug Abuse.
(Statistical Series 1, No. 6), Annual
Data 1986: Data From the Drug Abuse
Warning Network (DAWN). DHHS
Pub. No. (ADM) 87-1530. Rockville,
MD: The Institute, 1987a.
National Institute on Drug Abuse.
National Household Survey on Drug
Abuse: Population Estimates 1985.
DHHS Pub. No. (ADM) 87-1539.
Rockville, MD: The Institute, 1987b.
Negrete, J.C; Knapp, W.P.; Douglas,
D.E.; and Smith, WB. Cannabis af-
fects the severity of schizophrenic
symptoms: Results of a clinical
survey. Psychological Medicine,
16:515-520, 1986.
Nirenberg, T.D., and Maisto, S.A.,
eds. Developments in the Assessment
and Treatment of Addictive Behaviors.
Norwood, NJ: Ablex, 1987.
OTarrell, T.J.; Connors, G.J.; and
Upper, D. Addictive behaviors
among hospitalized psychiatric pa-
tients. Addictive Behaviors, 8:329-333,
1983.
Overall, J.E., and Gorham, D.R. The
Brief Psychiatric Rating Scale.
Psychological Reports, 18:799-812,
1962.
Parker, J.B.; Meiller, R.M.; and
Andrews, G.W. Major psychiatric
disorders masquerading as alcohol-
ism. Southern Medical Journal,
53:560-564, 1960.
VOL. 16, NO. 1, 1990
Pepper, B.; Kirshner, M.C.; and
Ryglewicz, H. The young adult
chronic patient: Overview of a
population. Hospital & Community
Psychiatry, 32:463-469, 1981.
Pokorny, A.D. The multiple readmis-
sion psychiatric patient. Psychiatric
Quarterly, 39:70-78, 1965.
Pope, H.G. Drug abuse and psycho-
pathology. New England Journal of
Medicine, 301:1341-1343, 1979.
Pulver, A.E.; Wolyniec, P.S.; Wagner,
M.G.; Moorman, C.C.; and
McGrath, J.A. An epidemiologic in-
vestigation of alcohol-dependent
schizophrenics. Ada Psychiatrica
Scandinavica, 79:603-612, 1989.
Richard, M.L.; Liskow, B.I.; and
Perry, P.J. Recent psychostimulant
use in hospitalized schizophrenics.
Journal of Clinical Psychiatry, 46:79-83, son, J,E., eds. Resource Materials for
1985.
Ritzier, B.A.; Strauss, J.S.; Vanord,
A.; and Kokes, R.F. Prognostic
implications of various drinking
patterns in psychiatric patients.
American Journal of Psychiatry,
134:546-549, 1977.
Robins, L.N.; Helzer, J.E.;
Croughan, J.; and Ratcliff, K.S.
National Institute of Mental Health
Diagnostic Interview Schedule.
Archives of General Psychiatry,
38:381-389, 1981.
Robinson, A.E., and Wolkind, S.N.
Amphetamine abuse amongst psy-
chiatric inpatients: The use of gas
chromotography. British Journal of
Psychiatry, 116:643-644, 1970.
Rockwell, D.A., and Ostwald, P.
Amphetamine use and abuse in
psychiatric patients. Archives of
General Psychiatry, 18:612-616, 1968. prehensive Psychiatry, 29:550-557,
Ross, H.E.; Glaser, F.B.; and Ger-
manson, T. The prevalence of
psychiatric disorders in patients
with alcohol and other drug prob-
lems. Archives of General Psychiatry, File, 1989.
45:1023-1031, 1988.
Roy, A. LSD and onset of schizo-
phrenia. Canadian Journal of
Psychiatry, 26:64-65, 1981.
Safer, DJ. Substance abuse by
young adult chronic patients.
53
Hospital & Community Psychiatry,
38:511-514, 1987.
Schneier, F.R., and Siris, S.G. A
review of psychoactive substance
use and abuse in schizophrenia: Pat-
terns of drug choice. Journal of
Nervous and Mental Disease,
175:641-650, 1987.
Schooler, N.R.; Hogarty, G.; and
Weissman, M. Social Adjustment
Scale II (SAS-II). In: Hargreaves,
W.A.; Atkisson, C.C.; and Soren-
Community Mental Health Program
Evaluators. Pub. No. (ADM) 79328.
Rockville, MD: National Institute of
Mental Health, 1979.
Schuckit, M.A. Alcoholism and
other psychiatric disorders. Hospital
& Community Psychiatry,
34:1022-1026, 1983.
Schuckit, M.A. Drug and Alcohol
Abuse. 3rd ed. New York: Plenum
Press, 1989.
Siris, S.G.; Harmon, G.K.; and
Endicott, J. Postpsychotic depressive
symptoms in hospitalized schizo-
phrenic patients. Archives of General
Psychiatry, 38:1122-1123, 1981.
Siris, S.G.; Kane, J.M.; Frechen, K.;
Sellow, A.P.; Mandeli, J.; and
Fasano-Dube, B. Histories of
substance abuse in patients with
post-psychotic depressions. Com-
1988.
Smith, W. A Profile of Health and
Disease in America: Mental Illness and
Substance Abuse. New York: Facts on
Solomon, P. Receipt of aftercare
services by problem types:
Psychiatric, psychiatric /substance
abuse and substance abuse.
Psychiatric Quarterly, 87:180-188,
1986.
54
Solomon, P., and Davis, J.M. The ef-
fects of alcohol abuse among the
new chronically mentally ill. Social
Work in Health Care, 11:65-74, 1986.
Spieger, D.L., and Harford, T.C. Ad-
dictive behaviors among youth. In:
Nirenberg, T.D., and Maisto, S.A.,
eds. Developments in the Assessment
and Treatment of Addictive Behaviors.
Norwood, NJ: Ablex Publishing
Company, 1987. pp. 305-318.
Spitzer, R.L.; Endicott, J.; and
Robins, E. Research Diagnostic
Criteria: Rationale and reliability.
Archives of General Psychiatry,
35:773-782, 1978.
Spitzer, R.L., and Fleiss, J.L. A re-
analysis of the reliability of
psychiatric diagnosis. British journal
of Psychiatry, 125:341-347, 1974.
Spitzer, R.L., and Williams, J.B.W.
Structured Clinical Interview for DSM-
IIIPsychotic Disorders Version. New
York: New York State Psychiatric In-
stitute, 1985.
Taylor, A.M., and Abrams, R. A
critique of the St. Louis psychiatric
research criteria for schizophrenia.
American Journal of Psychiatry,
132:1276-1280, 1975.
Thacore, V.R. Bhang psychosis.
British Journal of Psychiatry,
123:225-229, 1973.
Thacore, V.R., and Shukla, S.R.P.
Cannabis psychosis and paranoid
schizophrenia. Archives of General
Psychiatry, 33:383-386, 1976.
Treffert, D.A. Marijuana use in
schizophrenia: A clear hazard.
American Journal of Psychiatry,
135:1213-1215, 1978.
Trier, T.R., and Levy, R.J. Emergent,
urgent, and elective admissions.
Archives of General Psychiatry,
21:423-430, 1969.
Tsuang, M.T.; Simpson, J.C.; and
Kronfol, Z. Subtypes of drug abuse
with psychosis. Archives of General
Psychiatry, 39:141-147, 1982.
van Kammen, D.P.; Bunney, W.E.,
Jr.; Docherty, J.P.; Marder, S.R.;
Ebert, M.H.; Rosenblatt, J.E.; and
Rayner, J.N. d-Amphetamine-in-
duced heterogeneous changes in
psychotic behavior in schizophrenia.
American Journal of Psychiatry,
139:991-997, 1982.
van Kammen, D.P.; Rosen, J.; Peters,
J.; Fields, R.; and van Kammen,
W.B. Are there state-dependent
markers in schizophrenia? Psycho-
pharmacology Bulletin, 21:497-502,
1985.
Vardy, M.M., and Kay, S.R. LSD
psychosis or LSD-induced schizo-
phrenia. Archives of General
Psychiatry, 40:877-883, 1983.
Victor, M., and Hope, J.M. The
phenomenon of auditory hallucina-
tions in chronic alcoholism. Journal
of Nervous and Mental Disease,
126:451-481, 1958.
Walker, E.; Bettes, B.A.; Kain, E.L.;
and Harvey, P. Relationship of
gender and marital status with
symptomatology in psychiatric pa-
tients. Journal of Abnormal Psychology,
94:42-50, 1985.
West, A.P. Interaction of low-dose
amphetamine use with schizo-
phrenia in outpatients: Three case
reports. American Journal of
Psychiatry, 131:321-323, 1974.
Westermeyer, J., and Walzer, V.
Sociopathy and drug use in a young
psychiatric population. Diseases of
the Nervous System, 36:673-677, 1975.
Whitlock, F.A., and Lowrey, J.M.
Drug dependence in psychiatric pa-
tients. Medical Journal of Australia,
1:1157-1166, 1967
Wooley, L.F. The clinical effects of
benzedrine sulphate in mental pa-
tients with retarded activity.
Psychiatric Quarterly, 12:66-83, 1938.
SCHIZOPHRENIA BULLETIN
World Health Organization: Manual
of the International Classification of
Diseases, Injuries, and Causes of Death,
9th revision, clinical modification.
Geneva: WHO, 1979.
Zigler, E., and Levine, J. Premorbid
adjustment and paranoid-nonpara-
noid status in schizophrenia. Journal
of Abnormal Psychology, 82:189-199,
1973.
Acknowledgments
This research was supported by
USPHS grants MH-38636 and
MH-39998 from the National In-
stitute of Mental Health, and by a
grant from the Alliance for Research
in Schizophrenia and Depression
(NARSAD). Portions of the research
were presented at the 41st Institute
on Hospital and Community Psychi-
atry, Philadelphia, PA, October
15-19, 1989. The authors thank Erika
Brady, Leonard Burns, Pam Fawcett,
Liz Heiss, Maria Levitt, Tamara
Murdock, Colleen Quinn, Nina
Schooler, George Simpson, Ann
Tierny, Julie Wade, and John Wixted
for their assistance in completing
this project.
The Authors
Kim T. Mueser, Ph.D., and Hardeep
Singh, M.D., are Assistant Pro-
fessors of Psychiatry; Douglas F.
Levinson, M.D., and Randall L.
Morrison, Ph.D., are Associate Pro-
fessors of Psychiatry; Alan S.
Bellack, Ph.D., is Professor of
Psychiatry, Director of Adult
Psychology, and Director of the
Behavior Therapy Clinic; Kashinath
G. Yadalam, M.D., is Assistant Pro-
fessor of Psychiatry, and Director of
the Neuropsychiatry Clinic, Medical
VOL. 16, NO. 1, 1990
College of Pennsylvania at Eastern
Pennsylvania Psychiatric Institute,
Philadelphia, PA. Paul R. Yarnold,
Ph.D., is Assistant Professor of
Appendix 1. Epidemio-
logic Studies of Sub-
stance Abuse in
Schizophrenia:
Methodologic
Refinements
Medicine and Psychiatry, North-
western University Medical School,
and Adjunct Assistant Professor of
Psychology at the University of
1. Use standardized instruments to
diagnose psychiatric illness and
assess substance abuse. Psychiatric
diagnoses made without structured
clinical interviews (e.g., the Schedule
for Affective Disorders and Schizo-
phrenia) often have poor reliability,
even when modern diagnostic
criteria are used (e.g., DSM-III-R).
Patients' levels of current adaptive
functioning, response to psycho-
tropic medications, and interper-
sonal skill may bias the diagnostic
evaluations of treatment providers.
The use of standardized interview
instruments tends to reduce the ef-
fects of such biases. Similarly, the
use of standard instruments to
assess current and past substance
abuse (e.g., the Diagnostic Interview
Schedule) is necessary to provide
estimates of abuse prevalence that
can be compared across studies con-
ducted at different settings. Inter-
rater reliabilities for persons con-
ducting diagnostic and substance
abuse interviews need to be estab-
lished before evaluation of research
subjects and should be periodically
checked throughout the study. Few
previous studies have documented
the reliability of the assessments
performed.
2. Use multiple sources of informa-
tion to assess substance abuse. All
sources of information about sub-
stance abuse have questionable
reliability and validity. While
biological measures are the most ac-
curate, they are also the most
limited, since they cannot determine
social impairment resulting from
abuse. Interviews conducted with
55
Illinois, Chicago, IL. Kimmy Kee,
B.A., is a Graduate Student in
Clinical Psychology at Hahneman
University, Philadelphia, PA.
patients, treatment providers, and
significant others are most likely to
result in accurate assessments.
When information about substance
abuse is ambiguous, or when clear
discrepancies exist between different
informants, such patients should be
omitted from data analyses compar-
ing abusers of specific drugs with
nonabusers.
3. Evaluate both history of sub-
stance abuse and current abuse.
Most studies on schizophrenia sam-
ple a wide range of patients with
different ages and levels of chronici-
ty. The assessment of a history of
substance abuse may be more perti-
nent in evaluating whether
schizophrenia predisposes patients
to the abuse of particular drugs (or
whether abuse antedates the onset
of the illness) than current abuse
pattern. For example, fewer older
patients would be expected to be
currently abusing drugs, but their
history of drug abuse would be im-
portant in comparing prevalence
rates across diagnostic groups. Cur-
rent substance abuse may be more
reliably measured than history of
abuse, and it may be more relevant
to predicting outcome of the illness.
4. Assess substance abuse
prevalence in more than one
diagnostic group. The prevalence
estimates generated by any single
study will be limited in general-
izability to the specific setting and
demographic characteristics of the
sample. The extent and pattern of
substance abuse in schizophrenia
56
can best be evaluated by comparing
it with other patient or nonpatient
groups in the same or a similar set-
ting. Patient groups such as affective
or personality disorders are useful in
determining the specificity of a pat-
tern of substance abuse to schizo-
phrenia. The assessment of non-
patient controls in addition to
patient groups provides important
information on whether psychiatric
patients differ from others in their
vulnerability to substance abuse.
Controls must be examined who are
similar to patients in their demo-
graphic characteristics, minimizing
possible selection biases (e.g., seek-
ing "volunteers" through adver-
tisements may result in more literate
and motivated persons).
5. Match patient and nonpatient
groups on demographic variables.
Gender, age, race, and socio-
Minority Research
Training in
Psychiatry
economic status are related to
substance abuse in the general
population and psychiatric patients.
Different subject groups need to
have similar demographic character-
istics to allow conclusions about
diagnosis and substance abuse to be
drawn. Age and gender are par-
ticularly important to control for,
since groups of schizophrenic
patients are more likely to contain a
disproportionate number of young
males than is true in other diag-
nostic groups, and young males are
more prone to substance abuse. The
relations between demographic
factors and substance abuse should
be examined across and within diag-
nostic groups, to determine whether
similar persons are vulnerable in dif-
ferent groups.
6. Assess the abuse of specific
classes of substances and analyze
The American Psychiatric Associa-
tion (APA) is pleased to announce
the National Institute of Mental
Health's funding for the Minority
Research Training in Psychiatry Pro-
gram. This project was developed
from a recognition of the critical
need to train psychiatric researchers
for the future and to specifically
focus on the underrepresented pool
of talent represented by minorities
in the field of psychiatry.
The program is for medical
students, psychiatric residents, and
fellows. It will provide funding,
including a stipend and travel ex-
penses as well as other related train-
ing costs, for medical students to
have an elective or summer experi-
ence working in a research-intensive
department of psychiatry. Funding
for similar experiences for psychi-
atric residents on an elective
SCHIZOPHRENIA BULLETIN
the data accordingly. While many
studies have assessed alcohol and
drug abuse separately, few have ex-
amined specific classes of substance
abuse (e.g., stimulants, sedatives),
and none has attempted to examine
the abuse of different drug classes
simultaneously in relation to demo-
graphic characteristics or history of
illness. Without knowledge of the
pattern of abuse of different drugs
among schizophrenic patients, it is
not possible to test alternative
hypotheses about self-medication or
the influence of abuse on the course
of illness. Although substance
abusers tend to use a variety of dif-
ferent drugs, differences in preva-
lence of abuse between drugs exist
(e.g., cannabis is abused more fre-
quently than other illicit drugs). Fur-
thermore, the impact of continued
substance abuse on schizophrenia
can only be determined by examin-
ing individual types of substances.
basis will also be provided. In addi-
tion, there are opportunities for 1-
or 2-year postresidency fellowships
for research training in psychiatry.
The program also includes
development of a comprehensive
plan for identifying medical
students and residents who have an
interest in psychiatric research and
linking them with advisors and
mentors to counsel them about a
career in research. Combined with
the work experience, this will pro-
vide opportunity and incentive for
the medical students and residents
to pursue the option of research as
a career.
For further information about the
Minority Research Training in
Psychiatry Program, call Harold
Alan Pincus, M.D., or Jeanne
Spurlock, M.D., at the American
Psychiatric Association, 202-682-6238.